Fuzzy supervisory PI control of a binary distillation column via distributed control systems

In this paper, the fuzzy supervisory PI control is implemented via distributed control systems (DCS) to control a binary distillation column. The fuzzy c-mean clustering is used to identify the membership functions and fuzzy rules are determined using fuzzy gain scheduling technique. Thus, the need of heuristic method for designing fuzzy membership functions and rules from expert knowledge is omitted. Then, the fuzzy supervisors adapt the parameters of the PI controllers on line to maintain the top and the bottom temperature of the distillation column in spite of the changes in feed flow rate. Distributed control system instrumentation has had a major impact in allowing advanced control to be designed, implemented, operated and maintained.